The golden rule for digital therapeutics and connected medical devices

He who has the gold rules.   That’s all you need to know when it comes to privacy compliance.

In the past 5 years, a lot has happened in the digital health space. Venture funding in 2018 was close to $10BN and a lot of work is being done in the area of digital therapeutics and connected medical devices.

As our customers progress through their clinical trial journey to FDA clearance and post-marketing, we are frequently asked on how to achieve HIPAA compliance in an era of digital health apps, medical IoT and collection of RWD (real-world data) from patients.

I will try and help connected medical device engineering and regulatory managers make sense out of HIPAA and the HITECH Act (Health Information Technology for Economical and Clinical Health).

On January 25, 2013, the HIPAA Omnibus Rule was published in the Federal Register, which created the final modifications to the HIPAA privacy and security rule. You can see the source of the law here.

The HITECH Act created a supply chain trust model.

According to 45 CFR 164.502(e), the Privacy Rule applies only to covered entities (healthcare providers, health plans and healthcare clearinghouses). Going down the chain, covered entities have suppliers who are defined as BAS (business associates). A business associate is a supplier that creates, receives, maintains, or transmits protected health information on behalf of a covered entity or other business associates.

The HITECH Act requires suppliers in the chain of trust to comply with the Security Rule.   A medtech company and its’ cloud service providers, customer engagement service providers et al are all business associates.

The HITECH Act does not impose all Privacy Rule obligations upon a BA but:

1.BAs are subject to HIPAA penalties if they violate the required terms of their BA Agreement (BAA).

2.BAs may use or disclose PHI only in accordance with the required terms of its BAA

3.BAs may not use or disclose PHI in a manner that would violate the Privacy Rule if done by the CE

Down the supply chain and to the right

When we go downstream in the supply chain, the BAA becomes more and more restricted regarding permissible uses and disclosures.

For example, if a business associate agreement between a covered entity and a supplier does not permit the supplier to de-identify protected health information, then the business associate agreement between the supplier and a subcontractor (and the agreement between the subcontractor and another subcontractor) cannot permit the de-identification of protected health information. Such a use may be permissible if done by the covered entity, but is not permitted by the downstream suppliers in the supply chain, if it is not permitted by the covered entity’s business associate agreement with the contractor.

Concrete example of a digital therapeutic.

A physician (covered entity) prescribes a digital therapeutic app. The physician writes a script that is sent to a customer service center, which provides customer support to patients to download and use the app.

The healthcare provider will need a BA with the digital therapeutics company (or its customer service center that may be a separate business), who then has BAAs with other online suppliers for cloud and Braze customer engagement services. Graphically, the supply chain looks like this:

As we move down the supply chain and to the right, we see that the suppliers are providing specific and more restricted digital services.

Digital therapeutics HIPAA

 

The golden rule

Although a BA is a formal, regulatory requirement, it includes compliance with the HIPAA Security Rule and possible exposure to Privacy Rule disclosures. To a large degree, the Golden Rule applies – “He who has the gold rules”.   For early stage medtech and digital therapeutics companies, your customers have the gold. Do a good job on your homework on your security and privacy risk assessment.  Consider external threats as well as possible exploits and cascade attacks on your APIs.

Putting lipstick on the pig of electronic CRF?

esource for people in clinical trials

Good online systems do not use paper paradigms. In this post – I will try and entertain you with historical perspective and quantitative tools for choosing an EDC system for your medical device study.

Decades of common wisdom in clinical trials still hold to a paper-based data processing model. One of the popular EDC systems talks about the advantages of having online forms that look exactly like paper forms.  True – familiarity is a good thing, but on the other hand a digital UX has far more possibilities for user engagement and ease-of-use than paper.   So – it is, in a way admitting failure to provide a better UX and downgrading to paper.

We recently engaged with an Israeli medical device vendor who has an innovative device for helping solve a common medical indication for men over 50.

I won’t go into details.

If you are a guy over 50, you know what I mean.

If not, it doesn’t matter.

The client CEO was interested in an eCRF (electronic CRF – case report form) system.  eCRF is better than paper, but it is, at the end of the day just a paper form in an electronic format.

I was having a lot of trouble trying to understand the CEO’s business requirements.   My attempts to steer the conversation to a discussion of how to obtain fast data for his clinical trial and reduce his time to FDA submission fell on deaf ears. A follow up conversation and demo of Flaskdata with the clinical and regulatory manager focused more on reports and how to manage queries.  Queries are a vestige from the paper CRF period, where a study monitor would visit the research site once/month, compare the paper source with the electronic data entry and raise queries or discrepancies.

In order to put this process into a historical context, let’s compare accounting systems from the late 70s, early 80s to an eCRF system.

Accounting versus eCRF

Feature Accounting circa 1970 eCRF circa 2019
Input data Paper JV – journal voucher Paper source
Data entry Data entry to a 2-sided accounting system Data entry to an eCRF
Data processing A batch job, processes punch-card data entry and produces a data entry report and data error report Site coordinators enter data to a Web app 1-3 days after the patient visit. Data entry errors or invalid data create data validation queries which are ignored until the study monitor visit a month later
Exception reporting Data error report – with non-numeric or invalid dates Queries
Management reports Trial balance

Profit and loss

Cash flow

Bean counters of CRF/items

What is profit and loss?

What does a cash flow model have to do with clinical trials?

 

 

Cost justification and TCO for medical device EDC systems

My first recommendation would be don’t buy an EDC system just because its cheap. Charging $100-300/month for a data entry application is not be a reason to give someone money.  As a client of ours once said – “I know I can use Google Forms for data entry and its free but Google Forms does not have an audit trial so  Google Forms is not an option for clinical trials”.

As a rule-of-thumb, a good EDC system for medical device studies should include audit trails and a clinical cash flow report (the flow of patients in and out, the flow of data items in and out).   The EDC should also be able to produce a clinical Profit and loss statement, showing you how well you are doing on your primary and secondary efficacy and safety end points. A well-designed and well-implemented EDC should include a robust data model for testing the primary endpoint and collecting safety data.    At the minimum, a solid design and implementation will cost at least $10,000.  Over 10 months, that’s a starting cost of $1000/month.   As Robert Heinlein said – “There is no such thing as a free lunch”.

Your decision to buy EDC should be based on an economic breakeven point.   One breakeven method is based on cost reduction in site monitoring.  Assume the EDC system costs $4000/month (weighted cost including implementation) and assuming a monthly site visit costs $800/day, then your EDC system must be able to save 5 site visit a month and assure protocol compliance. This places an upper bound on the price you can pay.

This is albeit problematic for small, 1 site studies which often use DIY implementations.  Just remember, that ignoring the implementation cost, does not make the product cheaper. In other words calculate your TCO (Total cost of ownership).

Or as one wise man one said – I’m too poor to buy a cheap car.

Israel Biomed 2019-the high-social, low stress STEM conference

Impressions from Biomed 2019 in Tel Aviv

This week was the annual 3 day Biomed/MIXiii (I have no idea what MIXiii means btw) conference in Tel Aviv.  The organizers also billed it as the “18th National Life Science and Technology Week” (which I also do not know what that means). This was a particular difficult time for a conference of medical device and pharma in Tel Aviv since it coincided with the Eurovision 2019 activities – and the traffic was tough.

There were a huge number of lectures and participants from all over the world and I suppose from that perspective, the conference is a success and tribute to the burgeoning Israeli biomed industry.  Forbes calls Biomed “The High-Paying, Low-Stress STEM Job You Probably Haven’t Considered”.  I think that this is probably a good description for the conference – high participation but low stress.

My colleagues and I come to the conference to network, schmooze, meet customers and suppliers.  It’s a good opportunity to take a few meetings, say hi to friends and hustle for new business.  Having said that, I did meet a few really interesting companies:

RCRI – is a Minneapolis MN based medical device CRO.  I met Todd Anderson and his boss Lisa Olson and pitched our approach for fast data in clinical trials to assure high levels of patient compliance to the protocol and submit faster to FDA.    Todd and Lisa get it and they were open about the CRO business model being more people-hours not speed.     They seemed genuinely interested in what we are working on but its hard to tell with Americans.

Docdok Health – is a startup founded by Yves Nordman, who is a Swiss MD living in Carmiel.  It’s a doctor-patient communications platform beginning to branch out into Post-marketing studies with RWD.    We shared demos and it seems that there is synergy between our regulatory platform and their post-marketing work.

Resbiomed – met Alex Angelov, the CEO.  Alex is leading a consortium including Flaskdata, Carl Zeiss, Collplant, PreciseBio and Pluristem for a Horizon2020 submission for an amazing project for an implant to the cornea.  Dan Peres from Pluristem got us together.   Cheer for us!

BSP Medical and ICB (Israel China Biotech investment) – my buddy Hadas Kligman literally took me by hand to visit to Yehuda Bruner and Andrew Zhang and I did my 60s elevator pitch on getting medical device companies to FDA/CFDA 6-12 months faster.   We agree to talk after the conference.

Butterfly Medical –  I met Idan Geva, the CEO last year at Biomed – we ate lunch at the same table.  I pitched him but he was uninterested – they were using EDC2Go – and he didn’t want to hear other options.     At the Minnesota pavilion talking to Todd Anderson from RCRI,   Idan shows up and looks at me and says “Heah – Hi Danny – I left a contact me request on your web site yesterday and no one got back to me. I said shame on us.  He says – he was referred to us by someone from Florida who used to use Medidata.  I asked where/who? was it Miami?  He says yeah it was Miami and checks his phone – says its someone from Precision Clinical Research that are using Flaskdata and recommended.    (Precision is one of our customer’s Miami sites).  I asked what happened to EDC2Go – he said well you know – they are end of life (I think this means the end of low-cost EDC) and we are now entering questionnaires manually on paper and it is driving us crazy.   He said – can you stick around and give us a demo at 15:00?  I said sure.  We met at 15:00 by the bar upstairs in the David Intercontinental and I demoed the system – he said “Show me the Forms designer”. I showed him.  He says “show me how CRC enters data” – I showed him.  He says “Show me how to extract data” – I showed him.  I think he actually did not believe how fast the Extract to CSV process was and asked me twice if that was the data.  In the end – the format of Mac Numbers was a bit strange for him. I showed him a quick presentation – and he saw that Serenno is a customer – and says – “Heah Tomer is a neighbor of ours in the incubator in Yokneam”.    He asked how much and I said $2K for a basic onboarding package and $1500 / month.  Or $10K and we will build the CRF (their CRF is super simple btw).  He wanted a discount, being Israeli.  I said – “lets meet with your clinical person and get her to buy-in to the solution.  If she buys in – you and I can talk business but before that, there is no point horse-trading.

Count the probabilities of this happening and you will see that it is an impossible event.

Thursday I went back to demo Todd and meet Dr Yael Hayun from Syqe Medical. Yael is one of the most impressive people I’ve met in a long time. She is an MD from Hadassah and one of the movers and shakers in LogicBio Therapeutics.    After we chatted – I told her that Syqe is lucky to have her onboard.   I did our Today is about Speed presentation and a short demo. She was suitably impressed and then mentioned they had met with a Danish EDC company called Smart Trial – which turns out is yet another low-cost eCRF provider.   I said look – eCRF is like 10% of the solution you need – in the case of Syqe, you have a digital inhaler and with cannabis, you are going to have a lot of concerns about patient compliance.

This is what we do – fast data collection from patients, investigators and digital inhalers and automated deviation detection and response.

On the way back – huge traffic from Eurovision.   Didn’t hear a single lecture but the meetings and people were outstanding.

 

Living off generic solutions developed in the past

I recently read some posts on Fred Wilson’s blog and it was impressive that he writes every day.

I’ve fallen into the trap of collecting raw material and then waiting to find time to write a 2000-word essay on some topic of importance to me.   But, I think it was Steve Jobs who said the best time to do anything was 20 years ago and failing that – best time is now.  So now – I will start writing every day and attempt to write on topics of interest to my customers, not me.

We are working on automating patient compliance in medical device clinical trials.   Patient compliance is critical for the success of medical device studies.

When we mean success – we mean proving or disproving the scientific hypothesis of the study.  Efficacy – is the device an effective treatment for the indication?

Safety – is the device safe for patients?

When we say patient compliance automation we mean the combination of 4 things which depend on each other:

1.Reinforcing patient compliance to the protocol – for example reporting on time and taking the treatment on time.  AI-based reinforcement uses data from the patient’s behavior and similar behavior to keep the patient on track without driving him crazy with text or push messaging.

2.Automated monitoring of compliance – using clinical measures and the treatment schedule for the study.    An example of a clinical measure is the number of capsules a patient took. An example of treatment schedule is taking the capsules every day before 12.

The output of automated monitoring is real-time alerts and compliance trends to the study team.

3. Automate patient compliance reinforcement using and adaptive control process that takes fresh data from the alerts to make decisions on how to reinforce the patient and keep them on track.

4.In order to automate monitoring and do AI-based reinforcement of patient compliance, you need fresh and up-to-date data.

There is a lot of work being done by startups like Medable, Litmus Health and Flaskdata.io (disclaimer – I am the founder of Flaskdata.io) but it’s a drop in the ocean of 24,000 new clinical trials every year.

Fundamentally – the problem is that the clinical trials industry uses generic solutions developed 40 years ago to assure quality of data-entry from paper forms.

The generic solution used today involves waiting 1-3 days for site data collection to the EDC, and 4-6 weeks for a site visit and SDV and then another 1-12 weeks for a central monitoring operation in your CRO to decide that there was a protocol violation.

You don’t have to be a PhD data scientist to understand that you cannot assure patient compliance to the clinical protocol with 12-week-old data.

The only explanation for using 40-year-old generic solutions is that the CRO business model is based on maximizing billable hours instead of maximizing patient compliance.

It seems that if you want to achieve real-time detection and response and AI-based patient compliance reinforcement, you have to disrupt the CRO business model first.

Perverse incentives

The perverse incentive for the high costs of medical devices and delay to market

The CRO outsourcing model and high US hospital prices result in higher total CRO profits via higher costs to companies developing innovative medical devices.   These costs are passed down to consumers after FDA clearance.

We’ll take a look at the cost dynamics of medical device clinical trials and the clinical trial value chain.

We’ll then consider an alternative business model that changes the way medical device sponsors conduct clinical trials, reduce their costs by 70-80% and shortens time to FDA submission.

The high costs of US hospitals

By 2000, the US spent more on healthcare than any other country, whether measured per capita or a percentage of GDP.

U.S. per capita health spending was $4,631 in 2000, an increase of 6.3 percent over 1999. 4 The U.S. level was 44 percent higher than Switzerland’s, the country with the next-highest expenditure per capita; 83 percent higher than neighboring Canada; and 134 percent higher than the OECD median of $1,983. 5

It’s the prices, stupid.

In 2011, the US Affordable Care Act set a requirement for MLR (Medical Loss Ratio) that insurers must spend 80-85% of revenue on medical services.    This reduced insurer margins, and drove up hospital prices to make up for lower margin.

The CRO business model

CROS (clinical research organizations) are outsourcing businesses that provide an array of services for clinical trial management and monitoring, reporting and regulatory submission.   For medical device studies, CROS employ 2 basic outsourcing models, people sourcing and functional sourcing. In people out-sourcing, the medical device company is responsible for managing contractors. In functional outsourcing, the company may buy a set of functions, for example study monitoring and medical writing.

Neither CRO model has an explicit incentive to complete a study faster since that would reduce outsourcing revenue for the CRO. The more time a CRO spends on monitoring, site visits, SDV and study closeout, the more revenue it generates.

A medical device sponsor may elect to do it himself which shifts the CRO cost to an internal headcount cost supplemented with additional costs for consultants with risk and time delays by not having the CRO expertise and infrastructure. There is tacitly no free lunch, as we will discuss later in this article.

The result is a perverse incentive for delay and higher costs to bring innovative medical devices to market.

The CRO business model combined with higher hospital prices drive higher total profits via higher costs to customers. The higher cost of innovative medical devices is then passed down to consumers (patients) after FDA clearance.

Consumer value chains

A consumer value chain looks generically like this:

Suppliers -> Distributers -> Consumers

By the early 90’s, the PC industry led by Intel and Microsoft used a 2-tier value chain:

MSFT->Distis->Resellers->Customers.

Resellers were further segmented according the customer size and industry segment – Retail, Large accounts, SMB and VARS (value-added-resellers) selling their own products and services to a particular industry vertical.   The PC industry value-chain model left Microsoft with 50% of the SRP (suggested retail price) and delivered products to customers that were 45-50% less than SRP, leaving the channel with 0-5%.

The channel was forced to implement extremely efficient operations and systems and sell value-added services and products in order to survive.

By the new millennium, Apple introduced a 1-tier model with a user-experience designed and controlled by Apple.

The Apple 1-tier channel looks like this:

Apple->Apple Stores->Consumers

Eventually the Apple channel model broadened to include a 2-tier model similar to PC industry:

Apple->Distis->Retail->Consumers

By the mid-2000s, Amazon AWS (and generally the entire cloud service / SaaS industry) evolved the channel model to 0-tiers with a direct subscription and delivery model.

AWS->Consumers

As AWS grew and introduced spot pricing, an aggregation sub-market developed, looking extremely similar to movie and TV distribution models.

AWS->Aggregators->Consumers

AWS also became a distribution channel for other cloud products similar to content distribution (Think Netflix).

Third-party products->AWS->Consumers

Outstanding user-experience and aggregation are the hallmarks of companies like Airbnb, Netflix and Uber.    

The common thread is that AWS and Netflix deliver a digital product end-to-end, whereas Airbnb and Uber aggregate trusted suppliers inside the Airbnb and Uber brand environment and provide an outstanding and uniform user experience to all the consumers.  This is in contrast to the variegate user experience a customer got from the 90’s Microsoft channel. There are great resellers and terrible resellers.

We will return to user experience and aggregation later.

The medical device clinical trial value chain

The first published RCT (randomized clinical trial) in medicine appeared in the 1948 paper entitled Streptomycin treatment of pulmonary tuberculosis.

The clinical trial value chain for medical devices looks strange once after the historical perspective of how Intel, Microsoft, Amazon and Netflix evolved their value chains.

The medical device clinical trial value chain has 3 tiers with patients that are both suppliers and consumers.

Patients->Hospitals->CROS->Medical device companies->Patients

A dystopian user experience

Little has changed in the past 71 years regarding clinical trials.    Clinical trials and hospital operations now have a plethora of complex expensive, difficult-to-use IT with a value chain that provides a dystopian user experience for hospitals, patients and medical device companies.

HCOs (healthcare operators) rely on data collection technology procured by companies running clinical research (sponsors and CROs). This creates a number of inefficiencies:

1 – HCO staff are faced with a variety of systems on a study by study basis. This results in a large amount of time spent learning new systems, staff frustration and increased mistakes. This is passed on in costs and time to sponsors after CRO markup.

2 – The industry is trending towards the use of eSource and EMR to EDC data transfer. eSource/ePRO tools need to be integrated into the patient care process. Integration of EMR with EDC becomes logistically difficult due to the number of EDC vendors on the market (around 50 established companies).

3 – Siloed data collection in hospitals with subsequent manual data re-entry results in large monitoring budgets for Source Data Verification, and delays caused by data entry errors and related query resolution. Delays can be on the order of weeks and months.

4 – Use of multiple disconnected clinical systems in the hospital creates a threat surface of vendor risk, interface vulnerabilities and regulatory exposure.

Losing focus on patients

One of the consequences of the 3-tier medical device value chain is loss of focus on the patient user experience.  Upstream and to the left, patients are ‘subjects’ of the trial. The patient reported outcomes apps they use vary from study to study. Downstream and to the right (what FDA calls ‘post-marketing’), patients are consumers of the medical device and the real-world user experience is totally different than the UX in the study.   The real-world data of device efficacy and safety is disconnected from the clinical trial data of device efficacy and safety.

Clinical trial validation

Patient compliance is critical to clinical trial validation of medical device. Who owns patient compliance to the research protocol?  The medical device sponsor, the CRO, the hospital site or the subject?   The CRO may not collect a patient compliance metric since he outsources to the hospital. The hospital may not have the tools and the medical device company is outside the loop. My essay on determining when patient compliance is important in medical device trials goes into more detail on the problem of losing focus on the patient.

Vertical integration and aggregation

We previously made a qualitative claim that hospital site costs are high for medical device studies.  How high are they relative to consumer healthcare?

In a medical device trial recently done on the Flaskdata.io platform, the sponsor paid the hospital investigatory sites $700K for a 100 subject, 7 month multi-site study. (There were no medical imaging and blood test requirements).

In 2016, Medicare Advantage primary care spend was $83 PMPM (per member per month).      Let’s say that a premium service should cost $100 PMPM.    Let’s use that as a benchmark for the cost of processing a patient in a medical device trial.  Take this medical device Phase II medical device trial with 100 patients, running for 7 months:

That’s 100 x 7 x 100 = $70K for patients. Not $700K.

Perhaps the law of small numbers is killing us here.  The way to solve that is with aggregation and vertical integration. Let’s return to the medical device clinical trial value chain. As we can see, there are too many moving parts and a disconnect between patients in the clinical studies and consumers in the real world.

Patients->Hospitals->CROS->Medical device companies->Patients

One alternative is to integrate backwards and to the left.   This requires managing hospital site functions and to a certain degree is done in the SMO (site management organization model).

The other alternative is to integrate forward and to the right.   This is the path that Airbnb, Uber and Netflix took aggregating consumer demand with an outstanding user experience.  The aggregation gives Airbnb, Uber and Netflix buying power to the left, enabling them to choose the best and most cost-effective suppliers.

The value chain would then look like this:

Suppliers->Medical device companies->Patients

This is a model that we see increasingly with Israeli medical device vendors with limited budgets.   The Medical device company uses a cloud platform to collect digital feeds from investigators, patients and devices and automate monitoring for deviations. Focus on the patient user experience begins with design of the device and continues to post-marketing. Aggregation of patients enables purchasing power with suppliers – research sites, clinical consultants and study monitors.

Flaskdata - esource, ePro, patient compliance montoring,

Short-term versus long-term cost allocation

The reality is that using a technology platform for vertical integration is more expensive initially for implementation by the medical device company.   It should be.

Under-funding your infrastructure results in time delays and cost spikes to the medical device sponsor at the end of the study.

The current CRO methodology of study close-out at the end of a clinical trial lowers costs during the trial but creates an expensive catch-up process at the end of the study.

The catch-up process of identifying and closing discrepancies can take 2-6 months depending on the size and number of sites. The catch-up process is expensive, delaying submission to FDA and revenue since you have to deal with messy datasets.   The rule of thumb is that it costs 100X more to fix a defect after the product is manufactured than during the manufacturing process. This is true for clinical trials as well.   A real-time alert on treatment non-compliance during the study can be resolved in 5 minutes.   By waiting to the end of study it will take a day of work-flow, data clarifications and emails to the PI.

Summary

Vertical integration reduces costs and delay at study-end with continuous close. It is more expensive initially for the medical device company and it should be because it accelerates time to submission and reduces monitoring and close-out costs.

 

Patient compliance – the billion dollar question

The high failure rate of drug trials

The high failure rate of drugs in clinical trials, especially in the later stages of development, is a significant contributor to the costs and time associated with bringing new molecular entities to market. These costs, estimated to be in excess of $1.5 billion when capitalized over the ten to fifteen years required to develop a new chemical entity, are one of the principal drivers responsible for the ongoing retrenchment of the pharmaceutical industry. Therapeutic areas such as psychiatry, now deemed very high risk, have been widely downsized, if not abandoned entirely, by the pharmaceutical industry. The extent to which patient noncompliance has marred clinical research has in some cases been underestimated, and one step to improving the design of clinical trials may lie in better attempts to analyze patient compliance during drug testing and clinical development. Phil Skolnick, Opiant Pharmaceuticals The Secrets of a successful clinical trial, compliance, compliance, compliance.

Compliance, compliance, compliance

Compliance is considered to be key to success of a medical treatment plan. (1, 2, 3)

It is the “billion dollar question” in the pharma and medical device industry.

In home-use medical devices in particular and in chronic diseases in general – there is wide consensus that patient compliance is critical to the success of the clinical trial.   Our experience with Israeli innovative medical device vendors is that they understand the criticality of patient compliance. They “get it”.

However, as Skolnick et al note – patient compliance with the clinical protocol is often underestimated in drug trials.

There are 4 challenges for assuring patient compliance in medical device trials.

1. The first challenge is maintaining transparency.    An executive at IQVIA noted (in a personal conversation with me) that IQVIA does not calculate patient compliance metrics since they assume that patient compliance is the responsibility of the sites.    The sponsor relies on the CRO who does not collect the metrics who relies on the sites who do not share their data.

2. The second challenge is having common standard metrics of compliance. Site performance on patient compliance may vary but if sites do not share common metrics on their patients’ compliance, the CRO and the sponsor cannot measure the most critical success factor of the study.

3. The third challenge is timely data.   In the traditional clinical trial process, low-level data queries are resolved in the EDC but higher-level deviations often wait until study-closeout.  The ability of a study team to properly resolve thousands of patient compliance issues months (or even years) after the patient participated is limited to say the least

4. The final and fourth challenge is what happens after the clinical trial.  How do we take lessons learned from a controlled clinical trial and bring those lessons into evidence-based practice?

A general approach to measuring and sharing patient compliance metrics

A general approach to addressing these challenges should be based on standard metrics, fast data and active monitoring and reinforcement and reuse. 

1. Use standard metrics for treatment and patient reporting compliance. The metrics then become a transparent indicator of performance and a tool for improvement.

A simple metric of compliance might be a score based on patient reporting, treatment compliance and treatment violations. We may consider a threshold for each individual metric – for example a 3 strike rule like in baseball.

A more sophisticated measure of compliance might be similar to beta in capital market theory where you measure the ‘volatility’ of individual patient compliance compared to the study as a whole. (Beta is used in the capital asset pricing model, which calculates the expected return of an asset based on its beta and expected market returns or expected study returns in our case).

2. Fast data means automating for digital data collection from patients, connected medical devices and sites eliminating paper source and SDV for the core data related to treatment and safety endpoints.

3. Actively monitor and help patients sustain a desired state of compliance to the treatment protocol, both pharmacologic and non-pharmacologic. Not everything is about pill-counting. This can be done AI-based reminders using techniques of contextual bandits and decision trees.

4. Reuse clinical trial data and extract high quality training information that can be used for evidence-based practice.

Patient compliance teardown

Measures of patient compliance can be classified into 3 broad categories:

Patient reporting – i.e how well patient reports her own outcomes

1. Treatment compliance – how well the treatment conforms to the protocol in terms of dosing quantities and times of application. 2. Research suggests that professional patients may break the pill counting model

3. Patient violations – if the patient does something contrary to the protocol like taking a rescue medication before the migraine treatment

Confounding variables

Many heart failure patients are thought to be non-compliant with their treatment because of prior beliefs – believing that the study treatment would not help them. In the European COMET trial with over 3000 patients it was found that a Lack of belief in medication at the start of the study was a strong predictor of withdrawal from the trial (64% versus 6.8%; p < 0.0001). Those patients with very poor well-being and limited functional ability (classified as NYHA III–IV) at baseline significantly (p = 0.01) increased their belief in the regular cardiac medication but not in their study medication (4)

But numerous additional factors also contribute to patient non-compliance in clinical trials:  lack of home support, cognitive decline, adverse events, depression, poor attention span, multiple concomitant medications, difficulty swallowing large pills, difficult-to-use UI in medical devices and digital therapeutics and inconveniences of urinary frequency with diuretics for heart failure patients (for example).

It seems that we can identify 6 main confounding variables that influence compliance:

1. Patient beliefs – medication is useless, or this specific medication cannot help or this particular chronic condition is un-curable

2. Concerns about side effects – this holds for investigators and for patients and may account for levels of PI non-compliance.

3. Alert fatigue – patients can be overwhelmed by too many reminder message

4. Forgetfulness – old people or young persons. Shift workers.

5. Language –  the treatment instructions are in English but the patient only speaks Arabic.

6. Home support – patient lives alone or travels frequently or does not have strong support from a partner or parent for their chronic condition.

Summary

Flaskdata.io provides a HIPAA and GDPR-compliant cloud platform that unifies EDC, ePRO, eSource and connected medical devices with automated patient compliance monitoring. The latest version of Flaskdata.io provides standard compliance metrics of patient reporting and active messaging reminders to help keep patients on track.  Your users can subscribe to real-time alerts and you can share metrics with the entire team.

Contact Batya for a free demo and consult and learn how fast data, metrics and active reinforcement can help you save time and money on your next study.

References

1. Geriatr Nurs. 2010 Jul-Aug;31(4):290-8. Medication compliance is a partnership, medication compliance is not.
Gould E1, Mitty E. https://www.ncbi.nlm.nih.gov/pubmed/20682408

2. Depression Is a Risk Factor for Noncompliance With Medical Treatment: Meta-analysis of the Effects of Anxiety and Depression on Patient compliance. DiMatteo et al http://jamanetwork.com/journals/jamainternalmedicine/fullarticle/485411

3. Importance of medication compliance in cardiovascular disease and the value of once-daily treatment regimens. Frishman. https://www.ncbi.nlm.nih.gov/pubmed/17700384

4. Adherence and perception of medication in patients with chronic heart failure during a five-year randomised trial Ekman, Andersson et al. https://doi.org/10.1016/j.pec.2005.04.005

 

 

 

Teetering on the precipice of medical device/digital health clinical trials

Danny teeters on the edge of the precipice of privacy and security. Step on the brakes not on the gas and don’t look down. Take a 500m leap of faith into the chasm of medical device clinical trials. Validate digital therapeutics. Venture into uncharted territory of medical cannabis trials.

medical device clinical trials - leap of faith into it

At some stage in my “let’s do something different and risky” life after leaving the safety of Intel culture, I stumbled into cybersecurity.

Cybersecurity and privacy for medical devices

I started helping Israeli medical device and digital Health startups with privacy and security consulting. We built and analysed medical device threat models. The threat analysis approach succeeded in helping people improve their systems and privacy compliance.

Over time, the threat analysis methodology that was developed was adopted by thousands of security analysts globally – PTA Technologies.

Well-known digital health companies like Earlysense, Zebra Medical , Elminda, Dario Health, Tytocare, Intendu, as well as larger players like Biosense all worked with me on their HIPAA and FDA Cyber compliance posture at one point or another.

Compliance is a continuous process

I did not do this on my own. I owe these opportunities to my friend and colleague Mike Zeevi from Softquest Systems.

Over time, I figured out what works and how to comply with standards – HIPAA, FDA and GDPR. This came from real-life implementations and FDA submissions. I got hands-on in compliance audits with large US healthcare organisations like BC/BS Dignity Health.

Development practices for connected medical device and digital health apps

Many startups in the digital health and medical IoT space make 3 mistakes when engineering their systems.

1) First they Google. 2) Then they Guess. 3) Then they DIY when the Guesses Fail.

 

 

Some companies add an additional step: “Contract to a Software House that Talks Big” and then DIY or switch contractors.

This is a costly and risky pattern. As Jim McCarthy says –

More people have ascended bodily into heaven than have shipped great software on time.
– Jim McCarthy, Dynamics of Software Development by Jim McCarthy, Denis Gilbert

For Israeli digital health startups, there is an additional risk. This is the risk of not having an organisational memory. Youth has energy, hip viewpoints and updated expertise on latest technology. Who knew that a similar technology failed 30 years ago before you were born?

Build versus buy for digital health platforms

Digital health startups face 2 challenges. The first is an engineering challenge. The second is a validation challenge.

AWS cloud services have changed the way we engineer connected medical devices and digital health apps.

However, you need to factor in the cost and time requirements for a slew of additional activities. You need reliable DevOps, application integration, data integration, performance, configuration management, security, privacy, compliance and risk management.

The validation challenge is about clinical trials. About 4 years ago, we saw that our medical device customers wanted cheaper and faster ways to collect, monitor and analyse clinical trial data.

Building the product yourself and building a digital clinical trial systems is neither simple nor cheap. Resorting to paper studies to save money, turns short-term savings into long-term losses in time and data quality.

The solution – full-stack digital clinical trial platform

I joined forces with Jenya and we took a strange and wonderful decision to help Israeli medical device companies run clinical trials in the cloud.

This is what Flaskdata.io – patient compliance automation for medical device studies does. We provide a full-stack 21 CFR, HIPAA, GDPR compliant platform for collecting and monitoring data from investigators, patients and devices. Organisations like Theranica Therapeutics and Weizmann Institute all trust our platform for their human research. Today, Flaskdata.io helps site coordinators and clinical trial manager assure patient compliance using real-time alerts and trends at over 300 sites globally.

We work hard to bring modern technology to our customers instead of paper and save time and money.

Platform as a Service offerings like IBM Watson digital health has an amazing set of tools. You have to build your own product, integrate, test, secure, verify and validate.

By comparison, validated Software as a Service platform like Flaskdata.io enables you to get started immediately. You can design data collection using visual UI and integrate the open Flask API for medical devices. Check out our Swagger here.

There is a free tier that enables very early stage startups to start running pilots for free. And yes, we support, English, Hebrew and Chinese.

Give us a shot – you will not be sorry.

100X faster to deviation detection in medical device studies.

Automated Patient compliance deviation detection and response on the flaskdata.io platform for a connected medical device clinical trial is 100X faster than manual monitoring. Automated compliance monitoring analytics and real-time alerts let you focus your site monitoring visits on work with the PI and site coordinators to take total ownership and have the right training and tools to meet their patient recruitment and patient compliance goals.

When is patient compliance important in medical device clinical trials?

In this post, Danny Lieberman, founder of flaskdata.io , discusses when patient compliance is crucial for your medical device clinical trial and when patient compliance is a negligible factor to success of the study.

From adverse events to patient compliance

My original goal for Flaskdata.io was  to use machine learning to predict onset of adverse events during interventional medical device clinical trials. 

For that goal, we needed data, so we started by providing cloud EDC services for medical device clinical trials with high-touch personal service and attention to the quality of the data model.  Very quickly – it become apparent that we did not have enough data (and after 20 studies, hundreds of sites and thousands of patients), we still do not have enough data to predict adverse events.

However, after performing 6 digital clinical trials in 2 chronic disease indications (acute migraine and chronic constipation) we had an epiphany – “PATIENT COMPLIANCE IS KING”

Customers using the Flaskdata.io platform for digital clinical trials, collected data via the EDC from investigators, collected data from patients (via our ePRO app) and collected data from connected medical devices (via the Flaskdata.io medical device API). The evidence was overwhelming :

Patient compliance to the protocol is an acute issue and critical success factor to the success of a connected medical device clinical trial. 

Or is it.

Who owns patient compliance?  The sponsor, the CRO, the site or the subject?

This discovery of the importance of patient compliance made a profound impression on us because it came from customers and empirical data they collected in our EDC systems.   This impression would not change, although we began to hear dissenting opinion on the importance of, and responsibility for patient compliance in clinical trials.

Public discussion on trends in the clinical trials industry is heavily influenced by big pharmaceutical companies, big CROs like PPD and IQVia and a $70BN/year clinical operations services industry that deal largely with oncology and biotechnology.   When we spoke to biotech prospects about the ability of our digital clinical trials platform to accelerate time to regulatory submission and assure high levels of patient compliance – people smiled and said “Well automated compliance monitoring is an innovative approach, but in fact, patient compliance is not important to us”.   

We then spoke with the Israel country manager of one of top 3 global CROs – and they said “Interesting question.   We collect many clinical trials operations metrics, but patient compliance to the clinical protocol is not a metric we collect”.   I asked – “In that case, who is in charge of patient compliance? and the answer was – the sites”.  In this scheme of things, if patient compliance is not a CRO metric, then the sponsor has a blind-spot to what is possibly, the single most important factor to the success of his connected medical device clinical trial.  Or not.

After that, we spoke with the country manager of one of the top 3 pharmaceutical companies and Israel and he told us again “Patient compliance is a non-issue for us.  Patients come to the hospital and get treatment and there is no problem”. I asked him “What about psychiatry trials?”  He replied – “well yes, everyone knows that psychiatry trials have acute issues of patient compliance”

Hmm.

We then went back and did the most logical thing – searching in Google for “the critical success factors of clinical trials” and there are 290 million results and a ton of empirical evidence and academic and industry research on the importance of patient adherence in clinical trials.

And this vast body of empirical data is dealing primarily with drug trials, not medical device trials.   The VP Clinical of a gene therapy prospect (who had previously worked at a medical device company) told us that in gene therapy patient compliance is negligible while in medical device trials, patient compliance is acute.

Hmm again. So what does Google say?

The high failure rate of clinical trials has significant impact on providing potential curative treatments to patients in need….

One key factor that has been identified in the high failure rate of clinical trials is the adherence of patients participating in clinical trials to the dosing, treatment, and study procedures that are very carefully put in place in clinically rigorous protocols. Due to the rigor that is required in order to demonstrate an “effect” relative to a standard of care treatment, even a small deviation in medication adherence can result in a trial failing to meet its pre-specified clinical endpoint.

Additionally, the current nature of clinical studies include strict timelines and competition among sites to enroll eligible subjects which can many times result in the inclusion of subjects that are simply not “medication-compliant”. The issue of medication adherence is therefore one key factor sponsors should carefully look at monitoring closely when designing and planning the medical and operational oversight of their trials.

Unfortunately, the issue of medication adherence many times goes unmanaged and falls solely on site staff to oversee. As clinical sites are many times running multiple concurrent trials and are themselves pressed to remain productive, the one-on-one daily management of medication adherence of study subjects can many times be neglected. It is therefore in the best interest of the patients in need that sponsors look towards solutions that can help to support their clinical sites in providing additional resources to maintain close and frequent interactions with subjects enrolled in key studies. It simply is no longer sufficient to rely solely on very busy clinical practices to ensure successful adherence of patients in enrolled in trials.   

See Compliance – a key factor to a successful clinical trial.

I’m confused.  Is compliance the best interest of the patient or the best interest of the PI, or the sponsor or all of the above?  We know that the PI must monitor participants’ compliance with study requirements. Failure to monitor patients adequately can sabotage the entire study and damage the site’s reputation. 

CROS not collecting patient compliance metrics. Busy sites. Lack of tools. PIs who are generally not hands-on with the patients.   Sounds like a classic finger-pointing situation.  

We hear of the importance of site selection, but if patient compliance is not a CRO metric, then how do we measure site performance properly?  

The 4 quadrants of patient compliance

In fact, the question of HOW to measure the importance of compliance is intimately related to 4 factors – and interestingly enough is totally unrelated to the site or the PI.  The 4 factors of patient compliance are:

  1. How do you collect data?

  2. What is the indication?

  3. What does the product do?

  4. How involved is the patient in the treatment?

In order to understand why there is dissenting opinion on the topic of the importance of patient compliance – we can map life-science products into 4 quadrants:  (Patient-centric, Digital, Investigator-centric, Implanted).  The top right quadrant in green is a digital clinical trial for chronic disease, the top left is a traditional EDC operation with varying degrees of patient involvement, the bottom left is little patient involvement and EDC data collection from paper source and the bottom right is no patient involvement but with data collection from implanted devices (an interesting and extremely important use case in its own right).

The above picture tells the whole story.

Patient compliance in clinical trials is crucial in digital clinical trials and patient-centric trials using traditional EDC and patient reported outcomes.

In the end it is about the patient – not the PI, not the site operations team and their training, policies and procedures and not about the CRO.

But hey – this is something any sponsor worth their salt already knows.

Israeli Medical device innovation for high patient compliance

One of the most challenging problems in medical device clinical trials and in real-life is how to achieve high levels of patient compliance to the protocol.    
Automated patient compliance technology in medical device clinical trials is confronting CROs with an unpleasant status-quo of SDV as a low-value-add, high-cost, time-consuming activity for patient compliance assurance.  The approach that this company takes provides continuous patient monitoring without requiring patient compliance at all. 

EarlySense is an Israeli medical device that is based on a paddle placed under the patient’s mattress that continuously monitors patient movement, HR and RR trends.

The EarlySense device helps facilitate timely interventions for patients in non-ICU settings by adding a layer of care with continuous monitoring, drawing attention to those who may show early signs of deterioration and may require clinical intervention.

Since the EarlySense contact-free sensor (it looks like a small plate)  is placed under the mattress and there are no leads attached to the patient – there is no need for patient compliance.

EarlySense bedside unit

We spoke Dalia Argaman, VP Clinical & Regulatory at EarlySense to understand how to take medical devices from the design and engineering stages, navigate regulatory pathways and execute medical device clinical trials that receive FDA approval and save lives.

“I have a BSc in Chemistry and MSc in Chemical Physics and I was lucky enough to start my career immediately after completing my education. I joined Direx (a startup employing 6-7 people at that time) and became a part of developing an innovative medical device in the field of shockwave lithotripsy which is designed to break kidney stones without invasive procedure. Beforehand the standard of care was that patients with kidney stones had to undergo surgical procedure full of discomfort and further complications. I was the one who took the medical device through clinical trials  when the first prototype was created”.

It was back then that Dalia got introduced to the world of medical device clinical trials, submissions to regulatory authorities (FDA, CE, CFDA) and clinical data management.

Dalia has been on top of this innovative world through the 20 years of her career, working in several companies that develop different medical devices. Alongside success, she experienced a number of professional and personal challenges that shaped her career.

“You actually might be surprised to find out that the biggest challenge I had to overcome during my career was to try and balance family life with professional. I am a career person but I also have a family and I always had to juggle between being a mother, raising a family and moving forward on the career ladder. In addition, I think when you are a regulatory person, working with various people with different understanding, different backgrounds and fields and trying to get everyone at the same page is a huge professional challenge”, she said.

While working for Glucon (Developer of Non-Invasive Glucose Monitoring Devices in Israel) Dalia was involved in clinical trials with patients who had diabetes: that included children as well. She recalls that this experience was most memorable through her entire career as it required creating a medical device that would make patient compliance easy.

“People with diabetes are prone to numerous complications: patients have to monitor their blood glucose level constantly to avoid hyperglycaemia: the procedure is usually done by pricking the finger and drawing blood for analysis. You can only imagine how uncomfortable it is for young children or their parents who sometimes have to wake their child up several times a night to check it.  Being involved in a company that develops a device capable of making so many people’s lives easy is a subject of great pride for me”, she says.

Dalia Argaman is currently in charge of clinical regulatory affairs and quality assurance in Earlysense. The company develops contactless sensors that are placed under hospital bed mattresses and allow to monitor vital signs (heart rate, respiratory rate and other parameters) in a contactless way without making the patient feel uncomfortable.

“They can help the physicians, nurses to continuously supervise the patient and detect early signs of deteriorating in order to intervene early, thus reaching a better outcome”.   All of this is done using passive monitoring of the patient’s movement and without requiring patient compliance.

“There is currently a long delay between the time that a medical device is being developed by research and development teams, execution of medical device clinical trials, analysis of data received from clinical data management team, submission to FDA and the time that products get to the end users. The delay is connected with vigorous testing a product has to get through in order to be in compliance with standards and be approved. I think FDA understands well the importance of using automation to accelerate the process of executing clinical trials in order enable these medical devices to get to market and start saving lives sooner”.

Invisible gorillas and detection of adverse events in medical device trials

Weekly Episode #1 - Patients and study monitors are both people.

What is easier to detect in your study – Slow-moving or fast moving deviations?

This post considers human frailty and strengths.

We recently performed a retrospective study of the efficacy of  Flaskdata.io automated study monitoring in orthopedic trials. An important consideration was the ability to monitor patients who had received an implant and were on a long term follow-up program. Conceptually, monitoring small numbers of slow-moving, high-risk events is almost impossible to do manually since we miss a lot of what goes on around us, and we have no idea that we are missing so much. See the invisible gorilla experiment for an example.

One of patients in the study had received a spinal implant and was on a 6 month follow-up program dived into a pool to swim a few laps and died by drowning despite being a strong swimmer. Apparently, the pain caused by movement of the insert resulted  in loss of control and a severe adverse event. The patient had disregarded instructions regarding strenuous physical activity and the results were disastrous. 

It seems to me that better communications with the patients in the medical device study could have improved their level of awareness of safety and risk and perhaps avoided an unnecessary and tragic event.

Subjects and study monitors are both  people.

This might be a trivial observation but I am going to say it anyhow, because there are lessons to be learned by framing patients and monitors as people instead of investigation subjects and process managers. 

People are the specialists in their personal experience, the clinical operations team are the specialists in the clinical trial protocol. Let’s not forget that subjects and study monitors are both  people.

Relating to patients in a blinded study as subjects without feelings or experience is problematic. We can relate to patients in a personal way without breaking the double blinding and improve their therapeutic experience and their safety. 

We should relate to study monitors in a personal way as well, by providing them with great tools for remote monitoring and enable them to prioritize their time on important areas such as dosing violations and sites that need more training. We can use analytics of online data from the EDC, ePRO and eSource and connected medical devices in order to enhance and better utilize clinical operations teams’ expertise in process and procedure.

A ‘patient-centered’ approach to medical device clinical trials

In conditions such as Parkinsons Disease, support group meetings and online sharing are used to stay on top of medication, side effects, falls and general feeling of the patient even though the decisions on the treatment plan need to be made by an expert neurologist / principal investigator and oversight of protocol violations and adverse events is performed by the clinical operations team. There are many medical conditions where patients can benefit by taking a more involved role in the study. One common example is carpal tunnel syndrome. 

According to the findings of an August 3rd, 2011 issue of the Journal of Bone and Joint Surgery (JBJS), patients receiving treatment for carpal tunnel syndrome (CTS) prefer to play a more collaborative role when it comes to making decisions about their medical or surgical care. 

Treatment of carpal-tunnel syndrome which is very common and also extremely dependent upon patient behavior and compliance is a great example of the effectiveness of “shared decision-making, or collaborative, model” in medicine, in which the physician and patient make the decision together and exchange medical and other information related to the patient’s health.

As the article in JBJS concludes:

“This study shows the majority of patients wanted to share decision-making with their physicians, and patients should feel comfortable asking questions and expressing their preferences regarding care. Patient-centered care emphasizes the incorporation of individual styles of decision making to provide a more patient-centered consultation,” Dr. Gong added. 

In a ‘patient-centered’ approach to medical device clinical trials, patients’ cultural traditions, personal preferences and values, family situations, social circumstances and lifestyles are considered in the decision-making process.

Automated patient compliance monitoring with tools such as Flaskdata.io are a great way to create a feedback loop of medical device clinical data collection,  risk signatures improvement, detection of critical signals and communications of information to patients. Conversely, automated real-time patient compliance monitoring is a a great way of enhancing clinical operations team expertise.

Patients and study monitors are both people.